Models / Compare /DeepSeek V3.2 vs gpt-oss-120b
Head to head
DeepSeek V3.2 vs gpt-oss-120b
Compared on live catalog data — pricing synced daily, benchmark scores with provenance.
The short version
- gpt-oss-120b is cheaper to run hosted: $0.17/M output vs $0.40/M for DeepSeek V3.2 (2.4× difference at the cheapest tracked provider).
- DeepSeek V3.2 takes 164K tokens of context vs 131K for gpt-oss-120b.
- DeepSeek V3.2 scores 1469.5929 on Arena Coding vs 1390.1368 for gpt-oss-120b.
- DeepSeek V3.2 scores 1400.3738 on Arena Creative Writing vs 1277.4859 for gpt-oss-120b.
| Spec | ||
|---|---|---|
| Vendor | DeepSeek | OpenAI |
| Parameters | 685B / 37B | 120B |
| 164K | 131K | |
| Weights | OPEN | OPEN |
| License | MIT License | Apache License 2.0 |
| Modality | text->text | text->text |
| Released | Dec 1, 2025 | Aug 4, 2025 |
| 432.1 GBest | 75.9 GBest | |
| Cheapest hosted | $0.27 / $0.40 per 1M (Novita AI) | $0.037 / $0.17 per 1M (OpenRouter) |
| Tracked providers | 19 | 23 |
DeepSeek V3.2 — our take
DeepSeek V3.2 is a large downloadable text model with a permissive MIT licence and a 163,840-token request limit. Only 37 billion of its 685.4 billion total parameters are active per token, keeping inference costs down while measured coding performance is the clear peak in its benchmark profile.
full breakdown →gpt-oss-120b — our take
gpt-oss-120b is OpenAI's first downloadable model, released in 2025 with a permissive Apache licence and a 120.4-billion-parameter dense architecture. It offers broad measured coverage across chat, coding and creative tasks, though its scores vary sharply by category and it is text-only.
full breakdown →Need more columns? Use theinteractive comparison to add up to four models.